The cerebellum, a brain structure found in all vertebrates, is thought to serve many sensory-motor coordinating functions. In particular, decades of evidence suggest that the cerebellum guides motor learning based on sensory inputs. One laboratory example of motor learning thought to involve the cerebellum is a task in which a conditional stimulus, such as an audible tone, is associated with an unconditional stimulus, such as an airpuff delivered to the eye. Eventually the subject learns to blink in response to the tone played alone. For learning to occur the tone must come first, suggesting that the cerebellum can detect the order in which events occur. Under synaptic activity conditions thought to match the order-dependence of this task, one molecular event is elevation of calcium, a widely used signaling ion, inside cerebellar Purkinje neurons. Calcium, in turn, induces long-term depression, a form of synaptic plasticity at parallel fiber synapses to the Purkinje neuron that may underlie motor learning. This proposal will test the following ideas: (1) Conditional stimulus information coming in through parallel fiber synapses is encoded by the chemical messenger inositol-1,4,5-trisphosphate (IP3). (2) Unconditional stimulus information coming through the climbing fiber synapse is encoded by a small amount of priming calcium that enters during the action potential. (3) Synapses in the cerebellum generate large amounts of calcium release from internal stores when both IP3 and priming calcium are generated. (4) The order-dependence seen in calcium signaling is caused by dynamics in IP3, calcium, and calcium release mechanisms. This hypothesis will be tested using multiphoton laser scanning fluorescence microscopy, a technique that allows calcium signals to be measured at single synapses; and light-sensitive "caged" compounds, which can be used to control biochemical events in cerebellar Purkinje neurons. These experiments will provide crucial basic knowledge regarding the molecular events that occur on a time scale of milliseconds to seconds during synaptic plasticity and motor learning, and may eventually help in understanding defects in cerebellar function.